Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
IZA DP No. 830
Emigration from the UK,1870-1913 and 1950-1998
Timothy J. Hatton
DI
SC
US
SI
ON
PA
PE
R S
ER
IE
S
Forschungsinstitutzur Zukunft der ArbeitInstitute for the Studyof Labor
July 2003
Emigration from the UK, 1870-1913 and 1950-1998
Timothy J. Hatton University of Essex, Australian National University
and IZA Bonn
Discussion Paper No. 830 July 2003
IZA
P.O. Box 7240 D-53072 Bonn
Germany
Tel.: +49-228-3894-0 Fax: +49-228-3894-210
Email: [email protected]
This Discussion Paper is issued within the framework of IZA’s research area Mobility and Flexibility of Labor. Any opinions expressed here are those of the author(s) and not those of the institute. Research disseminated by IZA may include views on policy, but the institute itself takes no institutional policy positions. The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center and a place of communication between science, politics and business. IZA is an independent, nonprofit limited liability company (Gesellschaft mit beschränkter Haftung) supported by Deutsche Post World Net. The center is associated with the University of Bonn and offers a stimulating research environment through its research networks, research support, and visitors and doctoral programs. IZA engages in (i) original and internationally competitive research in all fields of labor economics, (ii) development of policy concepts, and (iii) dissemination of research results and concepts to the interested public. The current research program deals with (1) mobility and flexibility of labor, (2) internationalization of labor markets, (3) welfare state and labor market, (4) labor markets in transition countries, (5) the future of labor, (6) evaluation of labor market policies and projects and (7) general labor economics. IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available on the IZA website (www.iza.org) or directly from the author.
IZA Discussion Paper No. 830 July 2003
ABSTRACT
Emigration from the UK, 1870-1913 and 1950-1998∗
The international labour market has not been ‘globalised’ to the same degree over the last 40 years as have international markets for goods and capital. Immigration policies in developed economies clearly hinder the mobility of labour. But how much difference does it actually make? This paper compares emigration from Britain to four principal destinations in the era of free migration before 1914 with emigration to the same places since the 1960s. As the doors were kept open to British emigrants for longer than most, the ‘deglobalisation’ of British labour only dates from the 1960s. Since that time there has been a secular fall in British emigration, and this has been a major component in the transformation of the UK from a country of net emigration to one of net immigration. Before 1914 the economic and demographic forces that drove British emigration can be clearly identified. The same effects, applied to the later period, suggest that mass emigration from Britain should have continued until the early 1990s. But from the mid 1960s these influences became less powerful as they were increasingly inhibited by immigration policies in the principal destination countries. The decline in emigration is largely accounted for by shifts in policy, especially those that curtailed or abolished the preferences previously extended to settlers from the UK. JEL Classification: F2, J6, N3 Keywords: UK emigration, immigration policy Tim Hatton Economics Program Research School of Social Sciences Australian National University Canberra ACT 0200 Australia Tel.: +61 2 6125 4135 Fax: +61 2 6125 0182 Email: [email protected]
∗ I am grateful to the British Academy for support through a Research Readership. I also appreciate the useful comments of participants at the conference on “The Political Economy of Globalisation: Can the Past Inform the Present?” at Trinity College, Dublin, August 29-31, 2002.
1
Introduction It has become a commonplace to compare the depth and scope of globalisation
since the 1970s with the earlier era of open international markets from the middle of the
nineteenth century to 1913. Globalisation means that goods and factors of production are
internationally mobile and that, as a result, there is at least some degree of integration across
countries in these markets. A number of recent studies have suggested that rising
globalisation since the Second World War in goods and capital markets has brought the
developed world back to where it was a century ago (see the studies in Bordo et. al.
(2002)). By contrast (and despite much rhetoric in the press) international migration has
remained highly constrained (Chiswick and Hatton 2002). While many observers have
commented on the scale of international migration and the policies, that constrain it, there
have been few attempts to directly compare the determinants of international migration in
these two economic eras. With that as the broad agenda, this paper attempts to compare, as
directly as possible, the factors that drove emigration from Britain to the New World before
the First World War and after the Second World War.
British emigrants were among the pioneers of mass migration during the nineteenth
century. The chief destinations for British migrants were the United States, Canada,
Australia, New Zealand and what is now South Africa. After declining sharply during the
interwar period, mass emigration to the same destinations re-emerged in the 1950s and
1960s. Since that time British emigration has shrunk back to levels reminiscent of the
interwar period. This paper seeks to shed light on the following questions:
• Did the same economic variables drive British emigration in the two eras, and if so, how
did the magnitude of their effects differ?
• What was the role of immigration policy in the receiving countries and can its effects on
the numbers of migrants be identified?
• What combination of economic, demographic and policy variables accounts for the
striking de-globalisation of British labour since the 1960s?
The paper deals with these issues in four sections. The next section looks at the overall
magnitudes and trends in British emigration. This is followed by a discussion of the evolution
of immigration policies in the main receiving countries. The next section examines
econometric estimates of the determinants of emigration in the pre-1914 period. Those
2
estimates are then used to predict what might have happened from the 1960s to the 1990s
in the absence of policy intervention in the receiving countries. Emigration for this period is
then estimated directly to assess the effects of policy. The main conclusion is that most of the
fall in British emigration (and indeed much of the turnaround in net UK migration as a whole)
was due to immigration policy. In the absence of such policies, emigration of British citizens
would have been substantially higher, especially between 1971 and 1991.
Two Eras of Mass Migration
In the 60 years between 1853 and 1913 the passenger statistics record that a total
of nearly 13 million British citizens left the UK bound for extra-European ports, 10 million of
whom departed after 1870. These were passengers, chiefly travelling steerage, rather than
emigrants as such, and there was also a large return flow.1 Cumulative net passenger
movement after 1870 was nearly 6 million, equivalent to 13 percent of the population 1913.
About two three fifths of these emigrants were male, about three fifths were single, and
among the adults, more than four fifths were aged between 18 and 45.
The annual flow plotted in Figure 1 shows an upward trend in the total numbers, and
characteristic long swings with highs in the decade from 1865, in the 1880s, and in the
decade immediately before the First World War. The figure also shows net passenger
movements, possibly a better measure of emigration since it nets out return migrants and
non-migrant travellers. The absolute numbers are lower but the profile of net migration is
very similar to that of gross migration, indicating that fluctuations in the net numbers are
driven chiefly by the gross flows rather than by return migration. Although there is an upward
trend in the overall number of UK emigrants, there is no obvious long run trend in emigration
per thousand of the population. This ranged from lows of less than four per thousand in the
late 1870s and the late 1890s per thousand in to highs of around 8 per thousand in the early
1880s and in1910-12. In the 1850s and 1960s a large share of emigrants came from
Ireland, then part of the UK, but Irish emigration declined both absolutely and as a share of
1 British citizens (as distinct from aliens) in principle included those from anywhere in the British Empire although in practice the vast bulk were from Britain and Ireland. The figures were originally recorded by the Board of Trade. In 1912 migrants were distinguished from passengers for those travelling to the
3
the Irish population. This declining Irish contribution is evident in the diminishing gap
between emigration from the UK and emigration from Great Britain (England, Scotland and
Wales).
Most of the British emigrants travelled to one of four main destinations: the United
States, Canada, Australia and New Zealand. Together these account for 88 percent of the
gross outflow of UK citizens between 1871 and 1913 and for 93 percent of the net outflow.
As Figure 2 shows, the dominant destination among these was the United States, accounting
for 62 percent of the four-country gross outflow and for 55 percent of the overall total. Not
surprisingly, emigration to the US accounts for much of the year-to-year fluctuations in total
emigration observed in Figure 1. Emigration to Canada was much smaller and followed a
pattern of fluctuations somewhat similar to the US up to the turn of the century. There
followed a spectacular surge, that saw the numbers heading for Canada surpass those
bound for the US in the five years before 1914. For Australia and New Zealand (which are
combined in the Board of Trade’s statistics) the pattern of fluctuations was somewhat
different with a notable increase in 1910-13.
After a period of very low emigration in the troubled interwar years British
emigration revived in the 1950s and 1960s. For the years up to 1963 the UK statistics
record the flows of Commonwealth citizens travelling by long sea routes to ports outside
Europe and the Mediterranean, after which the statistics derive from the International
Passenger Survey, which also include travel by air. Both sets of statistics define migrants as
those travelling for an intended stay of at least a year, having spent at least a year at the
origin. In Figure 3 the gross outflow of Commonwealth citizens up to 1963 has been
adjusted upwards to allow for the increasing frequency of air travel, although no adjustment
has been made to the net figures.2 These suggest a total gross outflow from 1951 to 1998 of
7.3 million to non-European destinations, or 12.2 percent of the UK population in 1998, but
a much smaller net outflow of 1.7 million or just under 3 percent of the 1998 population. In
part this reflects growing in-migration of non-UK Commonwealth citizens.
USA, Canada and Australia/New Zealand in the last three quarters of 1912, emigrants amounted to 97 percent of total passengers (see Carrier and Jeffrey, 1953, p. 96) 2 The number travelling by air is assumed to increase linearly from 3,000 in 1949 to 50,000 in 1963. The adjustment for 1963 is based on the difference between emigration by long sea routes and an estimate for total emigration based on the pilot for the International Passenger Survey undertaken in 1963 (reported in the Board of Trade Journal, 1964).
4
From 1964 when UK citizens can be separately identified the pattern of fluctuations
is very similar to that for Commonwealth citizens. Both series show very strong downward
trends in gross and net migration from the mid -1960s. Between 1966-70 and 1994-98
gross emigration of UK citizens fell from an average of 199 thousand per annum to 76
thousand and the net figure fell from 124 thousand per annum to a mere 10 thousand. Most
of the gross emigration of UK citizens continued to be directed to the four traditional
destinations. From 1966 onwards they account for 62 percent of gross emigration and 86
percent of net emigration to non-European destinations. They also account for most of the
decline in the emigration of UK citizens between 1966-70 and 1994-98. Across these years
gross emigration fell by 123 thousand per annum while the four-country total fell by 100
thousand per annum, and the net figures fell by 114 and 92 thousand respectively. These
figures can also be compared with the overall net migration balance that includes foreign
citizens and also includes migration to and from Europe. The overall balance shifted from an
average net outward flow of 75 thousand per annum in 1966-70 to an average net inward
flow of 73 thousand in 1994-8. Thus the decline in the emigration of UK citizens to the four
traditional destinations accounts for nearly two thirds of the change in the balance that has
transformed the UK from a country of net emigration to a country of net immigration over
the last forty years.
Figure 4 shows the profiles of annual gross and net emigration to the four principal
destinations. Whereas in the period before 1914 the United States was the leading
destination, in the early postwar period it was Australia, which from 1966, accounted for 53
percent of the four-country gross emigration total. The steep decline in gross emigration to
Australia in the mid -1970s was followed by much lower average levels, with peaks in the
early 1980s and in the late 1980s and early 1990s. The profile for Canada shows a sharp
rise and decline in the 1960s followed by a series of fluctuations on a downward trend.
Emigration to the United States exhibits milder fluctuations in the 1960s and it is the only
country for which the total numbers of emigrants (either net or gross) increased between the
early 1970s and the late 1990s. By contrast the pattern of emigration to New Zealand
shows an upturn in the early 1970s followed by a gradual decline that resulted in net
immigration of UK citizens in the early 1990s. These patterns are closely replicated in the
immigration data of the receiving countries. Many observers would argue that these profiles
5
owe much to the changing immigration policy stance in the respective countries, a review of
which follows.
Immigration Policy in Two Eras.
Before the First World War there was essentially free migration for British citizens to
the main New World destinations. In the British Dominions there was effectively no
distinction between the native-born and those born in Britain and there was positive
encouragement for immigrants from Britain and Ireland. In Australia and New Zealand this
took the form of assisted passages that were allocated to prospective emigrants by
emigration agents operating in Britain. By contrast, Canadian recruitment schemes were
modest (and specifically targeted at prospective farmers) and there were no subsidies for
emigration to the United States.
In Australia, subsidies for free migrants to New South Wales began in 1832. Similar
schemes were adopted by other Australian colonies but they were gradually abandoned
between the early 1870s and the late 1880s. Subsidies were revived again from 1907 and
reached new peaks in 1910-1913.3 In New Zealand an energetic recruitment drive was
initiated in 1871 following the Immigration and Public Works Act of 1870. This scheme was
wound down at the end of the 1880s but revived again from 1906 and between 1908 and
1914 almost half of immigrants travelled on assisted passages (see Borrie, 1991). This
assistance covered only about half of the cost, which meant that a passage to the antipodes
was, for most of the period, still significantly more expensive than a passage across the
Atlantic.4
The pre-1914 regime was revived in the early post-1945 years. Although
restrictions on immigration had been imposed in the interim, policies that discriminated in
favour of British and Irish emigrants meant that there was still essentially free migration to the
3 Assisted emigration ceased in Victoria in 1873, in South Australia in 1886 in New South Wales in 1887 and in Tasmania in 1891. Such schemes were continued on a modest scale in Queensland and Western Australia. 4 In 1907 assisted male immigrants travelling third class to New Zealand paid £12 (2 berth cabin) or £10 (4 berth cabin) as compared with the total cost of £21 and £19 respectively. Rates charged to female domestic servants were £6.16s and £4.16s respectively. Similar rates prevailed in Australia until 1912 when minimum rates of £6 were agreed for farmers, farm hands, skilled artisans and nominated, assisted or indented male immigrants, and £3 for adult women. That compares with fares across the Atlantic that averaged £5.7s in 1910-1913 but that were mainly in the range of £2.10s -£4.10s during the previous 25 years (see Keeling , 1999).
6
three Dominions. In the United States the country-of-origin quotas, first introduced in 1921
and revised under the McCarran-Walter Act of 1952, made such generous visa allocations
to British emigrants that the quota was never filled. But from the 1960s restrictions grew
with a series of shifts in immigration policy. These policy changes can be characterised as
four types: (a) the removal or reduction of subsidies to British emigrants, (b) the introduction
of non-discriminatory policies that no longer favoured the British, (c) changes in global
immigration targets or quotas, and (d) the introduction of points schemes or other
mechanisms that shifted the balance of the selection criteria towards skills.
In the United States, the 1965 Amendments to the Immigration Act (effective June
1st 1968) abolished the country-of-origin quotas in favour of an overall quota for Eastern
Hemisphere immigrants and a preference system that allocated the majority of visas to
family-sponsored immigrants, although close relatives were exempted from the quota.
Immigration from the Eastern and Western hemispheres were merged into a global quota
from 1978. Under the 1990 Immigration Act (effective 1992) non-immediate relatives were
brought under the quota and a larger number of visas were allocated to skills-based
immigration under a revised preference system.5
In Canada, the preference given to immigrants from the UK, France and the USA
was abolished in 1962 and replaced with a scheme that allocated visas chiefly to sponsored
dependants, nominated relatives and independent migrants. The second and third of these
categories became subject to the points test that was established in 1967 (effective 1st
October). In addition to having relatives in Canada, points were awarded for age,
education, occupation and English or French language skills. From 1976 targets for total
immigration were set administratively, depending on economic conditions—a mechanism
that was used to effect a sharp reduction in immigration in 1982-5. The basic system was
modified in 1988 and 1993 further increasing the degree of skill selectivity such that, by
1994, nearly half of all immigrants were admitted principally on labour market
characteristics.6
5 There is a large literature on US immigration policy. Among those that report the details of this evolution, see Tomasi and Keely (1975), Briggs (1984) and DeLaet (2000). 6 For detailed discussions of the evolution and impact of Canadian immigration policy, see Hawkins (1989), Green (1994) and Green and Green (1999).
7
With a firmly established preference for British immigrants, Australia introduced the
UK Assisted Passage Scheme in 1947 offering passages to Australia for £10 per adult and
£5 for those aged 14-18. This was followed in 1957 by the ‘bring out a Briton’ campaign
with the result that between 1961/2 and 1971/2 only 10 percent of UK immigrants were
unassisted (Appleyard, 1988, p. 43). These schemes were reigned back in the 1970s in
parallel with the winding down of the white Australia policy that began in 1966. The final
abolition of discriminatory immigration policy took place under the Whitlam government in
1973 and became fully effective from 1975 (Hawkins, 1989; Jupp, 1991). At the same
time, administratively set immigration targets were reduced and new methods of selection
were sought. The first points system, somewhat similar to the Canadian System, was
launched in 1979 but was replaced by a new Migrant Assessment System from 1983.7
Over the following 15 years changes to this system involved introducing a business skills
stream and increasing the proportion of immigration subject to the points test (DIMA,
2000).
New Zealand was the last country to repeat the cycle in immigration policy followed
by Canada and Australia. As in Australia an Assisted Passage Scheme that offered £10
passages (but for key workers only) was introduced in 1947 and continued until 1975. The
preference for British and European immigrants was weakened from 1974 and abolished in
1987. The points system adopted in 1991 resembles those of Canada and Australia but
gives more weight to general skills rather than occupations in demand (Winkelmann, 1999).
As a result 65 percent of immigrants to New Zealand in the 1990s were points tested. Like
Australia the minimum threshold points score is adjusted depending on conditions in the
labour market.
7 The Numerically Weighted Multi-Factor Assessment System (NUMAS) that was introduced in 1979 has been characterised as a ‘halfway house” between the Canadian points system and the Structured Selection Assessment System (SAAS), largely based on interviews, that had existed since 1975. NUMAS came under severe criticism for giving too much weight to English language proficiency and thus undermining the shift towards non-discriminatory immigration policy (Hawkins, 1989, p. 142). The down-weighting of proficiency in English under the Migrant Assessment System in 1983 aroused even more controversy from those who still favoured a white Australia—see Blainey (1984).
8
Explaining UK Emigration 1872-1913
There is a large literature that concentrates on estimating the determinants of
migration flows from Europe to the New World in the late nineteenth century. Those studies
typically focus on bilateral flows from one origin to one destination or on aggregate
emigration from a particular source country or to a particular destination. They typically
specify emigration as depending on real wage rates and employment rates at the source and
the destination, sometimes including a limited number of other variables.8 The most recent
time series estimates for aggregate emigration from the UK Ireland and a number of other
countries follow this basic economic specification (Hatton and Williamson, 1998). This
model is derived for utility maximising individuals who are risk averse and who time their
migration in order to maximise the net present value of the move (see Hatton, 1995). A
version of that model is adopted here with the difference that migration streams to each of
the major destinations are combined in a pooled regression (rather than being combined in a
single aggregate series). The dependent variable is the Board of Trade’s gross or net
outflow of UK citizens per thousand of the UK population for the three destinations, the
United States, Canada and Australia and New Zealand combined. Details of the sources for
the explanatory variables can be found in the Appendix.
The result for gross emigration of UK citizens appears in Table 1. Two lags of the
dependent variable were found to be significant but lags of other variables, apart from those
included in the table, were not found to be important. The stock of previous migrants from
the UK at the destination concerned reflects the so called friends and relatives effect that
features so prominently in accounts that stress the cost-reducing and uncertainty-reducing
effects that migrant networks provide. Here, an addition of one thousand to the migrant
stock raises the annual flow of emigrants to that destination by 48 per annum in the short run
and by 94 per annum in the long run—a powerful effect consistent with those estimated for
aggregate emigration. The share of the UK population aged 20-34 is a proxy for the size of
the emigration intensive cohort—those for whom the net present value of migration is largest.
This effect must be multiplied by three to obtain the aggregate effect over the three
emigration streams. The coefficient implies that the emigration for this age group would be
8 For a survey of the earlier literature, see Gould (1979). Some of the more recent studies in this vein appear in Hatton and Williamson (eds.) (1994).
9
higher than that for other age groups by 13 per thousand in the short run and by 25 per
thousand in the long run.
The log of employment rates overseas and in the UK (lagged one year) have
powerful effects with signs that are opposite but have similar magnitudes. These results
strongly support the traditions in the literature that emphasises business cycle effects on the
timing of migration. In the long run across the three migration streams a one percent increase
in the destination employment rate increases the emigration rate by 0.9 per thousand in the
long run while a one percent increase in the UK employment rate reduces it by 0.7 per
thousand. The real wage ratio has an effect that is proportionately much smaller so that a
ten- percent rise in the destination to source wage ratio raises the combined gross emigration
rate by 0.9 per thousand.
A number of experiments were undertaken to capture the effects of subsidies for
emigrants to Australia and New Zealand in the 1870s and 1880s and again in the years
before 1914 by including dummy variables. The only one that was close to significance is for
the period 19010-13, reflecting the height of assisted passages to Australia. This is
consistent with Pope (1981) who finds a negative effect of passage costs on immigration to
Australia after the turn of the century. That effect is somewhat more significant in the second
column of the table, which is for emigration from Great Britain (excluding Ireland), a result
that makes sense since subsidies were targeted mainly to Britain. In other respects the
estimate for Great Britain is similar to that for the UK as a whole, suggesting that British and
Irish migration streams were guided by the same forces.9 The final column is for net UK
migration and again the pattern of coefficients is similar. Thus net emigration was driven by
very much the same forces as gross emigration.
The effects of the individual variables on the course of emigration can be assessed
by decomposing changes over time into the effects of different variables. Table 1 offers a
decomposition (for the three streams combined) using the long run coefficients to explain the
fall in the emigration rate between 1879-83 and 1894-8 and its subsequent rise between
1994-8 and 1909-13. For total UK emigration the effects of the migrant stock are negative
throughout as domestic population outstripped that of the stock. In the first period this was
9 Separate estimates for emigration from Ireland, in aggregate and to the USA, are presented in Hatton and Williamson (1998), Ch. 5.
10
offset by the growth in the share of the population most at risk. But more important than
these are the effects of destination country unemployment rates that reduced the overall
emigration rate by around three per thousand over the first period and then raised it by
about the same amount in the second period. Because these periods fall across the cycles of
unemployment in Britain, the domestic unemployment effects are much smaller. The wage
effects are more trended and these indicate that wage convergence reduced the emigration
rate by a little less than one per thousand in the first period. But from the mid 1890s to
1909-13 lagging domestic wage growth boosted gross emigration by more than one per
thousand of the population. The dummy for Australia/New Zealand also adds to the upsurge
in emigration in the years before 1914.
Predicting 1966-98 from the pre -1914 experience
Given the pre 1914 experience, what should we predict for the postwar period? As
we have seen, gross and net emigration to the four principal destinations fell secularly from
the mid-1960s. Is that what would have been predicted on the basis of the model estimated
for the pre-war period? Figure 5 shows, from 1966 onwards, total emigration of UK
citizens to the four destinations per thousand of the UK population. Gross emigration
declines from more than three per thousand in the late 1960s to around one per thousand in
the 1990s. This compares with an average for the pre-1914 period of 5.3 per thousand.
The dotted lines show what would be predicted for gross and net emigration using
the estimated coefficients from the pre 1914 period. These counterfactuals are obtained by
applying the coefficients in Table 1 (cols 1 and 3) to comparable annual time series (detailed
in the Appendix) for four destinations for 1966-98.10 The counterfactuals are adjusted to be
equal to the actuals in 1966. From that point, the predicted emigration rates fall slightly
before rising strongly in the 1980s. Then there is a sharp collapse in the early 1990s such
that the predicted emigration rates dip below the actuals from 1990. It is worth noting,
however, that the counterfactual predictions do not allow for the endogeneity of the migrant
10 The coefficients estimated for the late nineteenth century are all multiplied by 0.75, except for those on the lagged dependent variables. This allows for the fact that the pre-1914 coefficients are estimated over three emigration streams rather than four. The coefficients were then applied to the annual data for the four countries for 1966-98. These predictions were then summed over the four destinations and adjusted by an additive constant to be equal to the total in 1966.
11
stock. With higher levels of net emigration in the 1980s the stock would have fallen by less
and this would have attenuated the fall in emigration in the 1990s.
The long run coefficients estimated for the pre-1914 period are used to explain the
decline in predicted emigration in terms of the individual variables. As Table 3 shows,
between 1966-70 and 1980-4 the share of population aged 20-34 should have been
increasing the level of emigration, especially gross emigration. The emigrant stock and the
wage ratio had small negative effects while home and foreign employment rates had effects
that were almost exactly offsetting. Overall, gross emigration to the four countries is
predicted to rise by 2.3 per thousand whereas, in fact, it fell by 1.6 per thousand.
In the period from 1980-4 to 1994-8 both the migrant stock and the share of the
UK population aged 20-34 contributed to a predicted fall in gross and net emigration.
Employment rates had a small positive effect because the UK employment rate fell by
slightly more than did the average for the destination countries. But the most important factor
that reduced predicted emigration was the wage ratio, largely as a result of the sharp and
sustained rise in UK real hourly earnings in the early 1990s. Overall between 1966-70 and
1994-8 the prewar coefficients predict most of the decline in emigration but not its timing.
As Figure 5 showed, between 1971 and 1991 both gross and net emigration are massively
over predicted. This suggests that immigration policies in the receiving countries were
particularly important during those two decades.
Explaining Emigration 1966-98
The model used to explain emigration rates between 1966 and 1998 is similar in
spirit to that used for the pre-1914 period. The main difference is the inclusion of a range of
dummy variables that are intended to capture the shifts in immigration policies outlined
above. In the data for 1964-98 (see Appendix) Australia and New Zealand are separately
identified so that the model is estimated over four migration streams for the years 1996-98.
The model is estimated for UK citizens and for the total flow of UK and Commonwealth
citizens. The more inclusive definition is arguably closer to that for the pre-1914 period
when British and Dominion citizenship s were not distinct. But since the estimates that include
Commonwealth citizens are very similar to those for UK citizens alone, only the latter are
discussed.
12
The results in the first two columns of Table 4 bear some similarities with those
estimates for the late nineteenth century reported in Table 1. The first and second lags of the
dependent variable, though somewhat smaller than those estimated earlier, suggest a similar
pattern of dynamics. The effect of the stock of previous emigrants on gross emigration gives
a long-run coefficient such that increasing the emigrant stock by a thousand adds a further
33 per year to gross emigration. In the presence of family reunification policies, the stock
effect still matters, but the size of this effect is only one third as large as it was in the late
nineteenth century. The effect of the share aged 20-34 is also much smaller than it was in the
earlier era and the coefficients are much less significant. Overall an increase in the share aged
20-34 increases emigration in the long run by less than a tenth as much as it did before
1914.11
The effects of economic incentives as reflected in the coefficients on the employment
rates and the wage ratio are also very much smaller than those in the pre-war estimates. In
the long-run a rise in the employment rate across all destinations of one percentage point
raises gross emigration by a total of 0.15 per thousand. Although this reflects the fact that
immigration targets have often been responsive to employment conditions, the effect is still
only about one sixth of the size of that for the late nineteenth century. For the UK
employment rate no negative effect could be found in levels, but the change in the UK
employment rate does have a modest negative effect on both gross and net emigration.
Finally, the log of the wage ratio produces small and insignificant effects that are negative for
gross emigration and positive for net emigration. Thus, in the presence of destination country
immigration policies for most of the period, it comes as no surprise to find that the short- and
long run effects of economic variables are severely attenuated as compared to a free-
migration counterfactual.
Not surprisingly the dummies that represent country-specific policy shifts have
important effects particularly those occurring in the decade after 1966. The dummy for
Australia 1966-70 reflects the peak of activity in assisted emigration, and the effect of
winding down this policy was a sharp reduction in gross and net emigration to Australia. The
effect of the final abolition of the White Australia Policy in 1975 had an effect that further
11 This small effect is all the more surprising since age selectivity should be reinforced by points systems that give more weight to youth. For example the Australian system gives maximum age points
13
reduced emigration from the UK by a similar order of magnitude. But surprisingly, the initial
introduction of the Australian points system in 1979 had a positive effect that probably
reflects the easing of the immigration quota following the clampdown of the Whitlam years.
The effect of the introduction of the points system in Canada seems to have had a
large negative effect that probably reflects the tightening of immigration targets following the
unusually large immigration in the two years before 1968. By contrast, the effects of the
1965 Amendments to the US Immigration Act seem to have been only a very modest
reduction in emigration to the US. Changes in immigration policy in New Zealand and in the
United States in the late 1980s had positive but largely insignificant effects. This probably
reflects two things. First, as Figure 5 and Table 3 showed, pressure for emigration from
Britain had largely subsided by the early 1990s. And second, the shift towards skills-based
immigration policies probably favoured British immigrants on balance over those from other
potential source countries.12
In Table 5 the trends in the emigration of UK citizens are decomposed using the
long-run coefficients from equations 1 and 2 in Table 4. In the period from 1966-70 to
1980-4 the negative effects of falling employment rates overseas and the declining migrant
stock were partly offset by the falling employment rate in Britain and the rising population
share aged 20-34. But the key effects that reduced emigration were the tightening
immigration policies, especially in Canada and Australia. These shifts in immigration policy
account for 68 percent of the observed fall in gross emigration and for nearly all of the fall in
net emigration. In the period from the early 1980s to 1994-8 the fall in the emigrant stock
and the increasing UK employment rate added to the downward trend in emigration while
the policy variables evidently had small positive effects.
Over the whole period from 1966-70 to 1994-98 policy changes accounted for two
thirds of the fall in the four-country emigration rate of UK citizens, either gross or net. These
policy effects also account for about two fifths of the change in the overall balance of net
UK emigration across all citizenships and between all countries from 1.36 to –1.24 per
to those in the age range 18-29 and does not admit those over the age of 45. 12 Under the New Zealand points system English language proficiency is mandatory condition for skill-based immigration and it does not therefore contribute to the points score. For non-native speakers, minimum proficiency is measured as a score of 5 in each component of the IELTS test (Winklemann, 1999, p. 26).
14
thousand of the UK population. Allowing for the indirect effects of policy on the migrant
stock would make the contribution of immigration policy larger still.
Conclusion
Economic variables are uppermost in explaining fluctuations and trends in UK emigration
before the First World War. After the 1945 there was another emigration boom, to the
same destinations but on a slightly smaller scale. The evidence suggests that, had the
postwar flows been conditioned in exactly the same way as they were before 1914, the
surge of emigration would have continued until the early 1990s. Instead, net and gross
emigration of UK citizens declined secularly from the late 1960s, and it made a major
contribution to the shift in the overall UK balance of net migration. Shifts in immigration
policy in the principal destination countries had large negative effects on British emigration in
the 1970s and 1980s and, in the presence of these, the impact of economic variables on the
flows was much reduced. While policy was largely responsible for falling British emigration
in the 1970s and 1980s, the evidence suggests that by the 1990s the latent demand for
emigration had largely evaporated.
15
References Appleyard, R. T. (1988), Ten Pound Immigrants, London: Boxtree. Blainey, G. (1984), All for Australia, Sydney: Methuen Haynes. Bordo, M., Taylor, A. M. and Williamson, J. G. (eds.) (2002), Globalization and Histo ry, NBER (forthcoming). Borrie, W. D. (1991), Immigration to New Zealand, 1854-1938, Canberra: Highland
Press. Briggs, V. M. (1984), Immigration Policy and the American Labor Force, Baltimore:
Johns Hopkins University Press. Chiswick, B. R. and Hatton, T. J. (2002) “International Migration and the Integration
of Labor Markets,” in M. Bordo (et. al.). DeLaet. D. L. (2000), U.S. Immigration in an Age of Rights: Westport, Conn: Praeger. Department of Immigration and Multicultural Affairs (Australia) (2001),
Immigration: Federation to Century’s End, Canberra: Commonwealth of Australia.
Green, A. G. (1994), “A Comparison of Canadian and US Immigration Policy in the Twentieth Century,” in D. J. DeVoretz (ed.), Diminishing Returns: the Economics of Canada’s Recent Immigration Policy, Toronto: C. D. Howe Institute.
Green, A. G. and Green D. A. (1999), “The Economic Goals of Canada’s Immigration Policy: Past and Present,” Canadian Public Policy—Analyse de Politiques, 25, pp. 425-451.
Gould, J. D. (1979), “European Intercontinental Emigration, 1815-1914: Patterns and Causes,” Journal of European Economic History 9: pp. 41-112.
Hatton, T. J. (1995), “A Model of UK Emigration, 1870-1913,” Review of Economics and Statistics, 77, pp. 407-415.
Hatton, T. J. and Williamson, J. G. (eds.) (1994), Migration and the International Labor Market, 1850-1939, London: Routledge.
Hatton. T. J. and Williamson, J. G. (1998), The Age of Mass Migration: Causes and Economic Impact, New York: Oxford University Press.
Hawkins, F. (1989), Critical years in Immigration: Canada and Australia Compared,
Sydney: McGill-Queen’s University Press. Jupp, J. (1991), Australian Retrospectives: Immigration, Sydney: Sydney University
Press. Keeling, D. (1999), “ The Transport Revolution and Transatlantic Migration, 1850-
1914,” Research in Economic History 19, pp. 39-74. Pope, D. H. (1981), “Modelling the Peopling of Australia, 1900-1930,” Australian
Economic Papers 20, pp. 258-281. Tomasi, S. M. and Keely C. B. (1975), Whom Have We Welcomed? New York: Centre
for Migration Studies. Winkelmann, R. (1999), “Immigration Policies and their Impact: The Case of
Australia and New Zealand,” Bonn: IZA Discussion Paper No. 169.
16
Appendix: Data Sources A: 1870-1913 Emigration: Gross and net passenger movements from N.H. Carrier and J. R. Jeffrey, External Migration: A Study of the Available Statistics, 1815-1950 (General Register Office, London: HMSO, 1953), Table D/F/G (1). Inward movement for 1870-76 from I. Ferenczi and W. F. Willcox International Migrations, Vol. I (New York: NBER, 1929), Table XII, p. 640. Migrant stock: Benchmarks for the population born in the UK and in Great Britain, residing in the United States and Canada in census years from Carrier and Jeffrey (1953), Table 3. Australia from D. Lucas, “United Kingdom Born People in Australia,” in 1996 Census Community Profiles, (Canberra: Department of Immigration and Multicultural Affairs, 1996) p. 3. New Zealand from Census of Population (Wellington: Government Printer, various d ates). Annual migrant stock series were calculated by interpolating between census benchmarks using the relationship St = NM + dSt-1, where S is the stock, NM is net passenger movement detailed above, and d is a parameter calculated for each interval between censuses. Total population and share aged 20-34: For UK and Great Britain from B. R. Mitchell and P. Deane, Abstract of British Historical Statistics (Cambridge: Cambridge University Press, 1962), pp. 10-14. Share aged 20-34 linearly interpolated between census years. Unemployment rates: UK from G. R. Boyer and T. J. Hatton “A New Index of British Unemployment, 1970-1913,” Journal of Economic History 62 (2002) Appendix 1. United States from J. R. Vernon, “Unemployment Rates in Post-Bellum America, 1869-1899,” Journal of Macroeconomics 16 (1994), p. 710. For Canada an unemployment rate was constructed by regressing a non-linear transformation of the employment rate on the deviation of per capita real income from its logarithmic trend 1916-1926 and then extrapolating backwards. The employment rate was taken from W. Galenson and A. Zellner, “International Comparison of Unemployment Rates,” in National Bureau of Economic Research, The Measurement and Behaviour of Unemployment (Princeton: Princeton University Press, 1957), pp. 439-480, and real GNP per capita from M. C. Urquhart, “New Estimates of Gross National Product: Canada, 1870-1926,” in S. L. Engerman and L. E. Gallman (eds.), Long Term Factors in American Economic Growth (Chicago: National Bureau of Economic Research, 1986), pp. 9-94. Australia from N. G. Butlin, “An Index of Engineering Unemployment, 1852-1943,” Economic Record 22 (1946), pp. 241-260. This index was transformed using the relationship between unemployment in production industries and total unemployment estimated in Boyer and Hatton (2002). Real wage rates: Purchasing power parity adjusted unskilled wage rates for Great Britain, Ireland, USA, Canada and Australia from J. G. Williamson, “The Evolution of Global Labor Markets since 1830: Background Evidence and Hypotheses,” Explorations in Economic History 32, (1995), pp. 141-196. An index for the UK is calculated by giving a weight of 0.3 to Ireland and 0.7 to Great Britain. B: 1964-1998 Emigration: International Passenger Survey for gross and net migration; 1964-1973 from Board of Trade Journal (London: HMSO, various issues); 1974-1998 from International Migration (London: Office for National Statistics, various issues). Figures for UK citizens and Commonwealth citizens by destination for 1964-65 obtained by adjusting total migration by the shares of UK or Commonwealth citizens in total inward and outward flows for each destination country in 1966. Migrant stock: Decennial census benchmarks for the population born in the UK residing in the United States from Statistical Abstract of the United States (Washington: GPO, 1997). Canada (quinquennially from 1981) from Canada Yearbook (Ottawa: Statistics Canada, various years). Australia (quinquennially) from Lucas (2000), op.cit. New Zealand (quinquennially) from Census of Population (Wellington: Government Printer, various dates). The annual migrant stock was calculated, as for the pre-1914 period by using the net migrant flow to interpolate between census dates. Total population and share aged 20-34: Annual estimates of UK population, total and by age groups, Annual Abstract of Statistics, (London: HMSO, various issues). Unemployment rates: OECD Main Economic Indicators: Historical Statistics (Paris: OECD 1993) and subsequent annual issues of Main Economic Indicators. Real wage rates: 1975-98 hourly direct pay of production workers in manufacturing in national currency from US Bureau of Labor Statistics, “International Comparisons of Hourly Compensation Costs for Production Workers in Manufacturing, 1975-2000,” ftp://bls/gov/pub/ForeignLabor/supptabitxl. 1964-75 hourly earnings in manufacturing for UK, USA and Canada and consumer price index for UK, USA, Canada, Australia and New Zealand, 1964-98 from OECD (1993) op cit. and subsequent annual issues of Main Economic Indicators. Australia 1964-75 average minimum wage rates in manufacturing from Yearbook of the Commonwealth of Australia (Canberra: Australian Bureau of Statistics, various dates). New Zealand 1964-75 nominal
17
weekly wage rates from New Zealand Official Yearbook, (Wellington: Department of Statistics, 1982), p. 912.
18
Table 1: Explaining British Emigration, 1872-1913
(Pooled OLS regression with country fixed effects. Dependent variable: emigration per thousand of the UK population to the USA, Canada and Australia/New Zealand)
Independent variable UK Gross GB Net UK Net Emigration rate, t-1 0.78
(10.0) 0.81 (9.9)
0.78 (9.7)
Emigration rate, t-2 -0.29 (3.7)
-0.21 (2.6)
-0.32 (4.0)
Emigrant stock/home population 0.048 (5.2)
0.031 (4.2)
0.035 (3.7)
Share of population aged 20-34 12.62 (2.8)
9.61 (2.2)
4.50 (1.0)
Log destination employment rate 14.72 (6.9)
11.46 (6.1)
14.62 (6.1)
Log UK employment rate, t-1 -11.01 (5.0)
-9.87 (5.1)
-11.00 (4.5)
Log wage ratio (destination/home) 1.50 (3.6)
1.45 (3.9)
1.30 (3.0)
Dummy Australia and NZ, 1910-13 0.47 (1.8)
0.49 (2.1)
0.41 (1.5)
Canada 1.50 (3.3)
1.15 (2.9)
1.26 (2.4)
Australia and New Zealand 1.08 (2.9)
0.89 (2.8)
1.00 (2.4)
Constant -6.62 (1.9)
-5.32 (2.8)
-3.64 (1.8)
Adj. R2 0.93 0.90 0.84 D.W. 1.69 1.91 1.99 No. Obs. 126 126 126
19
Table 2: Decomposition of Trends in Emigration, 1879-83 to 1909-13
1879-83 to 1894-98 1894-98 to 1909-13 Effect of: Gross Net Gross Net Immigrant stock/UK population -1.13 -0.78 -1.57 -1.08 Share of UK population 20-34 1.65 0.53 -0.14 -0.04 Foreign employment rate -3.46 -2.76 3.17 2.98 UK employment rate 0.02 0.02 0.21 0.21 Wage ratio (UK/foreign) -0.85 -0.70 1.06 0.86 Australia dummy 0 0 0.92 0.76 Explained total -3.77 -3.69 3.65 3.69 Actual -3.39 -4.26 4.20 4.29
Table 3: Predicted Emigration Trends, 1966-70 to 1994-8 (Calculated using long-run coefficients in from Table 1)
1966-70 to 1980-4 1980-4 to 1994-8 Effect of: Gross Net Gross Net Immigrant stock/UK population -0.08 -0.05 -0.50 -0.35 Share of UK population 20-34 2.39 0.78 -1.30 -0.40 Foreign employment rate -4.46 -4.19 -0.11 -0.10 UK employment rate 4.44 4.19 0.41 0.38 Wage ratio (UK/foreign) -0.02 -0.01 -2.24 -1.84 Predicted total 2.27 0.72 -3.74 -2.34 Actual -1.60 -1.28 -0.27 -0.41
20
Table 4: Explaining British and Commonwealth Emigration, 1966-98 (Pooled OLS regression with country fixed effects. Dependent variable: emigration per
thousand of the population to the USA, Canada, Australia and New Zealand) UK Citizens UK and CW Citizens Independent variable Gross Net Gross Net Emigration rate, t-1 0.48
(6.5) 0.39 (5.0)
0.45 (6.2)
0.35 (4.5)
Emigration rate, t-2 -0.23 (3.4)
-0.20 (2.9)
-0.24 (3.6)
-0.19 (2.7)
Emigrant stock/home population 0.025 (2.9)
0.013 (1.4)
0.031 (1.4)
0.013 (1.2)
Share of population aged 20-34 0.89 (1.2)
1.89 (2.2)
1.07 (1.4)
3.40 (3.4)
Log destination employment rate 2.90 (5.4)
3.18 (5.0)
3.20 (5.6)
3.98 (5.5)
∆log UK employment rate -1.53 (2.3)
-1.76 (2.3)
-1.50 (2.2)
-2.47 (2.8)
Log wage ratio (destination/home) -0.14 (1.4)
0.06 (0.6)
-0.15 (1.5)
0.19 (1.6)
Australia 1964-70 0.39 (6.4)
0.40 (5.6)
0.39 (6.1)
0.42 (5.1)
Canada 1968-98 -0.37 (6.4)
-0.43 (6.0)
-0.41 (6.2)
-0.48 (5.8)
USA 1969-98 -0.04 (0.7)
-0.04 (0.1)
-0.04 (0.8)
-0.03 (0.5)
Australia 1975-98 -0.35 (5.7)
-0.34 (4.8)
-0.39 (6.1)
-0.38 (5.1)
Australia 1979-98 0.11 (2.5)
0.19 (3.7)
0.11 (2.2)
-0.48 (4.6)
New Zealand, 1988-98 0.08 (1.5)
0.09 (1.6)
0.09 (1.8)
0.20 (3.5)
New Zealand, 1992-98 0.05 (1.1)
0.10 (1.6)
0.06 (1.2)
0.11 (1.6)
USA, 1992-98 0.01 (0.02)
0.004 (0.1)
0.01 (0.3)
0.05 (0.8)
USA -0.18 (1.9)
-0.23 (2.0)
-0.26 (2.6)
-0.25 (1.9)
Canada 0.14 (1.9)
0.25 (2.7)
0.11 (1.4)
0.30 (2.8)
New Zealand -0.29 (2.7)
-0.29 (2.2)
0.32 (2.8)
0.30 (2.8)
Adj. R2 0.95 0.89 0.95 0.86 D.W. 2.04 2.00 2.34 1.88 No. Obs 132 132 132 132
21
Table 5: Decomposition of Trends in UK Emigration, 1966-70 to 1994-8 (For UK citizens using long run coefficients from Table 4)
1966-70 to 1980-4 1980-4 to 1994-8 Effect of: Gross Net Gross Net Immigrant stock/UK population -0.04 -0.02 -0.24 -0.12 Share of UK population 20-34 0.15 0.29 -0.08 -0.16 Foreign employment rate -0.80 -0.81 -0.02 -0.02 ∆UK employment rate 0.10 0.10 -0.19 -0.20 Wage ratio (UK/foreign) 0.00 0.00 0.19 -0.08 Australia dummies -0.84 -0.68 0.0 0.0 Canada dummy -0.49 -0.53 0.0 0.0 USA dummies -0.05 -0.05 0.01 0.01 New Zealand dummies 0.0 0.0 0.17 0.23 Policy effects -1.38 -1.26 0.18 0.24 Predicted total -1.97 -1.70 -0.16 -0.34 Actual -1.60 -1.28 -0.27 -0.41
22
Figure 1: Gross and Net Passenger Movement, UK citizens 1853-1913
0
50000
100000
150000
200000
250000
300000
350000
400000
450000
1853
1855
1857
1859
1861
1863
1865
1867
1869
1871
1873
1875
1877
1879
1881
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
Year
Ann
ual E
mig
ratio
n
UK Gross
GB Gross
UK NetGB Net
23
Figure 2a: Gross UK Migration to United States, Canada, Australia and New Zealand, 1870-1913
0
50000
100000
150000
200000
250000
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
Year
Nu
mb
er
USA
Canada
A/NZ
Figure 2b: Net UK Migration to United States, Canada, Australia and New Zealand, 1871-1913
-20000
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
Year
Nu
mb
er
USA
Canada A/NZ
24
Figure 3: Gross and Net Emigration, UK and Commonwealth citizens 1950-1998
-100000
-50000
0
50000
100000
150000
200000
250000
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
Year
Num
ber
Gross Emig: CW Citizens
Net Emig: CW Citizens Gross Emig: UK Citizens
Net Emig: UK Citizens
25
Figure 4a: Gross Migration to United States, Canada, Australia and New Zealand, UK Citizens
1963-98
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
Year
Nu
mb
er
USA
Australia
Canada
New Zealand
Figure 4b: Net Emigration to United States, Canada, Australia and New Zealand, UK Citizens
1963-1998
-10000
0
10000
20000
30000
40000
50000
60000
70000
80000
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
Year
Nu
mb
er
Australia
Canada
USA
New Zealand
26
Fig 5: Counterfactual UK Emigration to US Canada Australia and New Zealand, 1966-98
-1
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
Year
Em
igra
tio
n p
er 1
000
Po
pu
lati
on
Counterfactual Gross
Counterfactual Net
Actual Gross
Actual Net
IZA Discussion Papers No.
Author(s) Title
Area Date
816 T. T. Herbertsson J. M. Orszag
The Early Retirement Burden: Assessing the Costs of the Continued Prevalence of Early Retirement in OECD Countries
3 07/03
817 T. M. Andersen T. T. Herbertsson
Measuring Globalization 2 07/03
818 J. Pencavel
The Surprising Retreat of Union Britain 3 07/03
819 M. Beine F. Docquier H. Rapoport
Brain Drain and LDCs’ Growth: Winners and Losers
1 07/03
820 C. M. Cornwell K. H. Lee D. B. Mustard
The Effects of Merit-Based Financial Aid on Course Enrollment, Withdrawal and Completion in College
6 07/03
821 P. Carneiro J. J. Heckman
Human Capital Policy 6 07/03
822 D. Weichselbaumer R. Winter-Ebmer
The Effects of Competition and Equal Treatment Laws on the Gender Wage Differential
6 07/03
823 A. Filippin Discrimination and Workers' Expectations
5 07/03
824 A. Filippin Discrimination and Workers' Expectations: Experimental Evidence
5 07/03
825 A. Filippin A. Ichino
Gender Wage Gap in Expectations and Realizations
5 07/03
826 K. T. Hansen J. J. Heckman K. J. Mullen
The Effect of Schooling and Ability on Achievement Test Scores
6 07/03
827 H. Buddelmeyer E. Skoufias
An Evaluation of the Performance of Regression Discontinuity Design on PROGRESA
6 07/03
828 D. Checchi T. Jappelli
School Choice and Quality 3 07/03
829 J. J. Heckman X. Li
Selection Bias, Comparative Advantage and Heterogeneous Returns to Education: Evidence from China in 2000
6 07/03
830 T. J. Hatton Emigration from the UK, 1870-1913 and 1950-1998
1 07/03
An updated list of IZA Discussion Papers is available on the center‘s homepage www.iza.org.